The good professor's arguments are asinine and deadly wrong. Retranslated, "I see no reason why you should be concerned about the dangers of a so called "atomic explosion". With the tiny amount of U-235 you have managed to isolate, you have barely managed to demonstrate more than the slightest bit of warmth resulting from radioactive decay. I see no reason to believe your extraordinary claims that it will detonate in a flash with the energy equivalent to thousands of tons of explosives"
The evidence that superintelligent AI is better than our evidence for nuclear fission in the 1940s. We know we are intelligent. We know you can build an intelligent machine with 86 billion neurons * 1000 interconnects. We know that our current neural hardware is thousands of time slower than the clock rates of trivially constructible digital circuits.
Today, our current efforts are tiny. Most advanced artificial neural network models only use 1000 or so connections, not the trillions we know it actually takes. We don't have the patterning for arranging those trillions of connections properly. Similarly, if you isolated a gram of U-235 by itself, it would seem harmless.
The theories in the 1940s knew it wasn't harmless, that putting enough of it in one place would lead to a chain reaction that would create nuclear driven heat. They eventually built, at great expense, the first reactors to test this.
Similarly, we know that the rate we think limits how fast we can invent and prepare new technology. We know that technology vastly better than what we have now is possible - machines that are constructed with careful thought atom by atom, that can self replicate and rearrange matter at an atomic level...
The professor's wrong, and is the same blithe ignorance as stating that it's totally ok with slam together those pieces of purified U-235 with no safety measures.